Cutaneous melanoma is increasing in incidence faster than all other malignancies in the United States except for lung cancer in women. While most cases are discovered early when cure is likely, about 30% of patients will die from their disease. Patient prognosis is determined by primary tumor thickness, ulceration and the presence of regional lymph node metastases. Neither serologic nor molecular markers in the blood have been shown to be accurate to predict outcome. Elective lymph node dissection (ELND) was described over 100 years ago as a technique to identify lymph node metastases but this procedure remains controversial as the therapeutic value of this procedure is unproven. While the tumor status from the regional lymph nodes supersedes other prognostic factors, the cost and morbidity of ELND make it not justifiable for routine use. In 1990, lymphatic mapping and sentinel lymphadenectomy (LM/SL) was described as a minimally invasive alternative to ELND. LM/SL enables surgeons to identify a single lymph node (sentinel, SN) in direct connection with the primary melanoma. LM/SL has been demonstrated to be a highly accurate method for identifying occult lymph node metastases and has replaced ELND as a staging tool. LM/SL allows investigators a unique opportunity to examine the physical interaction of the primary tumor and SN. We hypothesize that the SN and primary melanoma are intimately related through a local regional immune response that is initiated by antigen-presenting dendritic cells. This interaction of primary skin tumor and regional lymph nodes is important for understanding the natural history and potentially the prognosis in individual patients. We have observed morphologic and phenotypic features of the SN that suggest the SN may be functionally different than non-SN from the same patient and that SN with metastases differ from similar SN without disease. While understanding the altered lymph node phenotype may lead to more individualized staging, methods to reverse these changes may be necessary for creating clinically relevant therapies. Our preliminary studies have shown that the morphologic features of the SN can be at least partially restored to that of non-SN through a single preoperative peritumoral injection of granulocyte-macrophage colony stimulating factor (rhGM-CSF). The goal of this proposal is to further characterize the morphologic and phenotypic features found in SN and determine the significance of these alterations. Our hypothesis is that the deficit of DC activation and presence of immunosuppressive type cytokines from the SN are important for the initiation of the metastatic cascade of melanoma, and dendritic cell agonists such as GM-CSF may prevent the metastatic cascade. Our aims are: 1) To determine if the morphologic and phenotypic differences in the SN and non-SN in melanoma are of prognostic importance and the relationship of these changes with the primary site, 2) To determine if the alterations in SN morphology are reversible, and 3) To determine if the restoration of the morphology or phenotype of sentinel lymph nodes results in diminished regional tumor burden.